In this section we will illustrate how an “and then”-prefaced turn continuation works as a same-speaker continuation in a post-gap position. Prior to Excerpt 3, one of the participants, Harry, has asked Tanja if she could briefly provide the background of the artist Frida Kahlo, whose artwork the participants are discussing. Just before the sequence in (3), the participants have established that Kahlo was injured in a traffic accident and at line one Tanja is explaining what kind of consequences the accident had.

First, Tanja states that Kahlo’s legs did not work properly. On line 3, she starts another TCU, where she elaborates that one of the legs was shorter than the other, after which she modifies the description by adding and specifying that Kahlo suffered from a more comprehensive bodily injury (line 6). Here Tanja uses the Swedish discourse particle alltså ‘I mean,’ which indicates a specification, correction or reformulation of something that was stated earlier (see Lindström, 2008:108). Simultaneously with Tanja’s telling about Frida Kahlo’s injuries, Raila asserts that the traffic accident affected Kahlo’s hip (lines 2, 4, and 7–8).

Up to this moment, Tanja has been accompanying her speech with so-called depicting (or iconic) hand gestures that have moved up and down her thighs, legs and shoulders (lines 1–6). These gestures, alongside language, function as a kind of representation of Kahlo’s injuries as she knows them (see Streeck, 2009: 9). At her specification in line six of the fact that Kahlo’s whole body was broken, she pauses her gestures letting her hands rest on the table before she starts another explanative TCU in line 9, where she says that Kahlo always wore a long skirt because of her broken body. This explanation is again accompanied by depicting hand gestures moving alongside the speaker’s legs (lines 9–13).

In line 10, Harry receipts that Tanja’s description refers to something he can see in the artwork as well. Thereafter Tanja continues her turn in line 11 with the consecutive conjunction så att ‘so that’ specifying that by wearing a long skirt Kahlo could avoid people seeing her feet. Here Tanja’s intonation contour is falling, which suggests that her turn is completed. Harry reacts with the change-of-state token aijaa ‘right’ (see Green-Vänttinen, 2001) followed by okej ‘okay’, i.e. he marks receipt and acceptance of new information. However, Tanja still adds the short coordinated phrasal unit å ben ‘and legs,’ also with a falling intonation, which now suggests a definitive closure of her contribution (line 13) and reestablishes a TRP that was already reached at line 11. The sense of completion is enhanced by the movement of Tanja’s hands to a rest position on the table during the gap that follows in line 14 (see Image 3.1).

The slight gap at line 14 following Tanja’s recompleted turn leaves a space for her to extend the contribution at line 15, beginning with the linking element å sen “and then.” This post-gap continuation takes a new direction in Tanja’s descriptive project: she is not speaking about Kahlo’s injured body but more generally about her appearance which she, after some hesitation, labels as strange. She then elaborates on this and highlights Kahlo’s eyebrows, which she refers to in German (Augenbrauen), directing her gaze to the other participants and asking for help with the Swedish word (va e dihär “what are these?”), and by pointing with her index fingers at her own eyebrows. On line 19 Raili fills in the Swedish word for ‘eyebrows,’ which Tanja repeats (in an incorrect plural form), concluding that Kahlo’s eyebrows were strong.

Tanja has had her hands on the table during the gap prior to the “and then”-prefaced continuation in line 14. While producing the connective å sen, her hands still lie on the table in a rest position (Image 3.1), but as Tanja after a micropause sets out with the turn continuation, the hand position changes so that the palms of her hands turn flat against the table, preparing to move again (Image 3.2 in line 16). Thus, å sen is accompanied by multiple semiotic resources that work together to mark an incipient move forwards in conversation (cf. Mondada, 2015). Furthermore, as Tanja progresses by focusing on Frida Kahlo’s eyebrows, she again uses depicting gestures that support her verbal description (Image 3.3), as she did in the preceding talk about Kahlo’s injuries. Thus, throughout Excerpt 3, Tanja accompanies her speech with hand gestures that mark junctions between beginnings and completions of turn-units as well as references to the world as she knows it (Frida Kahlo’s appearance and injuries) and as a means to localize aspects of the immediate environment and tying them to the linguistic form (the word for eyebrows).

Sequentially, the post-gap continuation is enabled by no uptake by the other participants, which results in a short gap. The conjunctional preface links back to the prior talk but also suggests the next relevant thing that can be said about the topic: the continuation gives a new direction to the speaker’s account of Frida Kahlo, embodying a shift from the description of her injuries to her looks. Gestures play an important role in the trajectory of the extended telling. Depicting and pointing gestures occur during the central parts of Tanja’s account while retraction to a rest position marks completion and a possible TRP. It is noteworthy that gesticulation is activated just after the production of the lexical linker å sen ‘and then’–a pattern we follow up in the following excerpts.

In the section above we showed how an “and then”-prefaced turn continuation worked in a post-gap environment. In this section we will show how a same-speaker continuation works when it occurs post-other talk. In Excerpt 4, the participants are discussing an abstract painting by the Japanese artist Yayoi Kusama. The painting consists of small dots in different white and greyish shades. Prior to the sequence in (4), the artwork has been assessed as boring by Jenny, a stance that is acknowledged but modified by Katarina, who thinks that the painting is neutral, yet pretty. Peter then begins to render what comes to his mind when looking at Kusama’s work.

Peter starts off by claiming no-knowledge with ja vet int vaffö men ‘I don’t know why but,’ accompanied with a slight head shake, when reporting what has made him think of a certain children’s fantasy fiction character while looking at the painting (line 1). After a small pause, he then continues with an explanative clause när här finns dehär ‘because here is this,’ with which he foregrounds an area in the painting. Simultaneously, with this verbal element he moves out from a bodily rest position and locates the relevant area in the artwork with his right arm and index finger (line one and Image 4.1 in line 3) describing it as a little darker than the rest of the painting, still circling around this area with his index finger (lines 3–5). He then retracts his arm so that he can rest his head against his hand and continues by mentioning that for some reason the Moomin character, The Groke, comes to his mind (lines 7–8 and Image 4.2). The coordinated verbal demonstratives (‘here,’ ‘this’) and pointing gestures enable Peter to establish a certain place in the picture (and draw the other participants’ attention to that place) before progressing with more elaborated thoughts about the painting (cf. Goodwin, 2017: 223–227; Heath and Luff, 2013).

Peter’s turn-ending in line eight can be understood as a local pragmatic completion point–signaled by complete syntax, falling intonation and a return to “a thinking rest position” (Image 4.2)–that is, “a point at which the speaker is projecting more talk, but at which another speaker might reasonably take a minimal turn” (Ford and Thompson, 1996: 150). Accordingly, Peter’s contribution evokes a change-of-state reaction from Jenny, aijaa (‘right’) followed by okej “okay,” who thus expresses some surprise. The other participants respond to Peter’s characterization with laughter (lines 9–11). In response to these receipts, Peter extends his turn in line 12 with a continuation prefaced by å sen, which here comes post-other talk (that consisted of minimal responses and laughter). With this expansion he gives a motivation of how his impression of The Groke could be understood. Peter’s account contains several discourse particles that relativize the description, such as liksom “like, sort of” and sådär “like, sort of,” or appeal to intersubjectivity (vet du “you know”).

The turn continuation beginning with å sen (lines 12–13) is accompanied by Peter’s move from a rest position to a gesture. He is now trying to explain the reason for his association of The Groke, swaying his fingers loosely in the air at the same time as he suggests that the area that he has pointed to looks a bit like mist (line 12, image 4.3). He then retracts his arm to the same kind of “thinking” rest position as he did at the previous completion at line 8 (Image 4.2) while continuing to describe in a little more detail how he pictures the Moomin character fuzzily in the misty, wintery background (see line 13).

Nevertheless, Peter’s multimodal explanation does not yield any more convinced responses from the other participants at this point. In overlap with Peter’s description both Jenny, Roni, and Katarina produce minimal responses, and after Peter has finalized his extended turn in line 13 with a falling intonation contour, Jenny responds with a hesitant mhm intressant ‘mhm interesting’ (line 17). In the following lines (19–21), Peter tries to convince the others again of what can be seen and imagined in the picture. With a lengthened nå (‘well’) in line 22, Jenny still does not display any great conviction, but Roni approves of Peter’s suggestion by responding sant (‘true’) in line 23, which Jenny then also registers with a quiet mm in line 24.

The post-other talk continuation in Excerpt 4 is designed interactionally as an elaboration of a point that was presented and projected in a previous turn by the speaker. At the intervening TRP, i.e. a local pragmatic completion point, the other participants take a recipient position and produce minimal responses that ratify Peter’s upcoming turn continuation. By expanding his turn with the additive-enumerating å sen, Peter offers an account of what he is experiencing in the artwork. The gestures at these kinds of motivating continuations are of a conceptualizing nature, where content is made sense of (cf. Streeck, 2009: 9, 160–171). In comparison, the continuation that was produced post-gap (see section Post-gap Continuation) added something new and was accomplished with more depicting gestures that aimed to represent the world as the participant knows or imagines it (cf. Streeck, 2009: 9). The division between these two kinds of gesture is not strict, and they are often lumped together and labeled as iconic or imagistic gestures (Streeck, 2009: 9; Kendon, 2004: 101–106). Yet, as the activity that is in focus for this study is task-based, where the participants describe, interpret, and assess artworks, it seems to be significant that the gestures also have a qualitative difference depending on whether the participants are describing something known to them or if they are trying to make sense of their observations. The general dynamics of the embodied cues around possible TRPs is nonetheless a trajectory starting from a rest position to gesture and back to a rest position, where retraction signifies possible completion.

The turn extensions that have been analyzed above were produced by the same speaker. In this section we will illustrate how the next speaker can make a turn entry through å sen and continue the thematic thread from the prior talk. In Excerpt 5, the participants are discussing the American artist Jean-Michel Basquiat’s naivistic painting Mona Lisa, which is painted in a quite childlike-manner entailing many symbols, like dollar signs. The participants discuss the painting and the representation of Mona Lisa and the connotations that the artwork evokes.

Prior to the excerpt, Lena and Amanda have noted that the face in the portrait looks distorted or like a man who is a heavy drinker. Amanda then makes a reference to a vivid story about the Norse god Thor (lines 1–3). Thor had flaming eyes when he was getting married and the portrait makes Amanda think about this connection. The other participants meet Amanda’s description with laughter and Jonna displays vague recognition with the particle a ‘ah’ (line 6). As a response to the amused reactions, Amanda at line seven reiterates her point that the person portrayed is somewhat reminiscent of Thor.

Amanda’s description is met with minimal responses in the lines that follow (8–9), from Lena with a hesitant hm and from Jonna with the receipt ha ‘right.’ A gap of 0.4 s follows, when everyone is looking at the artwork and a series of minimal response tokens, possibly signifying that the topic is ebbing away, are produced. At line 13, Lena takes the turn by introducing it with å sen, in so doing constructing it as an expansion of the collective interactional project on assessing the Basquiat painting. Through this continuation Lena brings in a new quality of the artwork that has nothing directly to do with the Norse god that Amanda had talked about. Instead, Lena points out the breasts’ appearance in the portrait and tries to find a suitable adjective to describe the visual impression.

During the pause in line 10 that precedes Lena’s next-turn continuation, the participants have retracted to bodily rest positions, leaning toward the table and keeping their arms and hands on the table. When Lena takes the turn at line 12, she moves her right hand over the image of the artwork to point at the new item she wants the others to focus on. At the demonstrative här ‘here’ she keeps her right index finger ready to point at both breasts in the picture (Image 5.1), thus, singling out one describable element from potential other ones. Lena’s pointing is done with an open palm, which gives the others a better visual access, but also presents the describable element as something worth examining (cf. Kendon, 2004: 2010–214). The pauses and hesitating sounds in lines 14–16 indicate that Lena is searching for a suitable descriptive label for the breasts, which is also the point where she retracts her arm to a rest position (line 15–16). Amanda makes a collaborative completion with the adjective glansiga ‘glossy (plural)’ in line 17, which Lena accepts with jå ‘yeah’ in line 19. Amanda then elaborates on this with the jocular question about whether the person in the portrait is wearing spandex (line 20), which invites laughter from the other participants. At lines 24–25 Emma adds an aspect she notices in the artwork (the flag of Finland), which is again met with laughter. Finally, in line 29 Lena responds with an alternative

to Amanda’s previous suggestion about spandex, eller som glänsande läder ‘or like shiny leather.’ This is met with some minimal responses that bring the descriptive sequence to a closure (lines 31–34). The participants have not really agreed on any specific description or interpretation of the painting; rather, they have introduced different noteworthy, more or less serious visual aspects of the artwork in a collective effort in trying to make sense of the unconventional portrait.

The connective å sen links Lena’s contribution to the prior talk, but also clearly moves in a new direction within the thematic progression. It brings in something that is important from the current speaker’s perspective and does not elaborate on or align with the prior speaker’s line of reasoning. It is possible that the additive meaning of “and then” makes both the speaker change and the shift of footing more subtle in the flow of interaction (cf. Stoenica et al., 2020), i.e. the local transition is masked as an expansion of the prior talk and as a collaborating effort in the face of it. Other-continuations are almost exclusively accompanied by a gesture that points out the new aspect that is brought into discussion. The embodied pointing is related to the verbal making of a point, tying gesture to language and the ongoing and situated meaning making (cf. Goodwin, 2017: ch. 15; Streeck, 2009: 9).

In our data, where the participants actively discuss and evaluate artworks, language is necessary for operations like description, interpretation, and other forms of reasoning. However, to make sense of the artworks, multimodal cues serve as resources in the collective, ongoing, and incrementally emergent arguing and sense-making. In our analysis in sections Post-gap Continuation–Other-Continuation, we have illustrated how different gestures and their trajectories work together with the verbal language to single out turn closings (retractions to a rest position), to localize and point out observables in the artworks (pointing gestures), to serve as reference to the experience that is known to the speaker (depicting gestures), and as a resource for making sense of experiences (conceptualizing gestures).

The TRPs that precede self-continuations, both post-gap and post-other talk, are indicated with multiple cues: complete syntactic gestalts, ⁶ falling intonation, and hand/arm and body movements that settle in a rest position. Following either no uptake (Excerpt 3) or ratifications of listenership (Excerpt 4), the speakers then produce a continuation prefaced with å sen ‘and then,’ where å sen initiates another multimodal sequence where talk and gestures work together to both depict and point out observables (Excerpt 3) as well as to support a line of reasoning (Excerpt 4). In both types of continuations (post-gap and post-other talk), the connective å sen serves as the linking element that enables expansion, and it is produced with the speaker still in a bodily rest position. Only after the verbal link is established do hands start to move and accompany the speech. Hence, it seems that gesture is not participating in making a link to the prior talk–that may be an abstract discursive-temporal task that is reserved for the connective element ⁷ –but belongs to the content part of the contribution. This embodied dynamics in the organization and achievement of self-continuations, sensitive to possible TCU completions and TRPs, can be schematized as follows:

In relation to the linking practices, something similar can be observed in the studies by Keevallik (2017) of dance classes, where conjunctions and prepositions link demonstrations of correct and incorrect performances in instructing activities. In Keevallik (2017) the subsequent embodied turn was often accomplished without speech, and some body orientation was observable when the linking conjunction was produced; however, the verbal link seemed necessary although the rest of the demonstration could be embodied. The activity at hand in our data is more language driven. Yet, as illustrated in our analysis, the participants also rely heavily on gestures, both as additional referential material for a description (esp. with L2 speakers) and for making sense of an experience (with L1 speakers).

Finally, other-continuations also show a clear pattern, where a pointing gesture typically occurs simultaneously with a verbal demonstrative. This is in line with previous studies that have established that pointing localizes observables in the surrounding environment and directs the participants’ attention to the next relevant thing to talk about (cf. Goodwin, 2017: 223–227; Mondada, 2016: 341–347, Mondada, 2014b; Kendon, 2004: 205–208). At the onset, the next speaker is–having listened to the prior turn–in a rest position. Similar to same-speaker continuations, the linking element is established first without a prominent embodied act. Only after the link å sen ‘and then’ is produced, does the speaker make an embodied move, stretching out an arm to point at the focused items in the image on the table in front of the speakers. In these “and then”-prefaced turn continuations, then, grammar in the form of the lexical linker precedes and paves the way for other multimodal resources to contribute to meaning-making in the content-wise more central and complex turn-parts (cf. Keevallik, 2017, Keevallik, 2013).

As the data contains recordings of conversations with both L1 and L2 speakers it deserves to be commented on whether we can see some differences in the deployment of turn-continuing practices in the respective groups. The L2 speakers in our data are at an advanced level and their proficiency in Swedish oral use is very good. When we have compared the use of å sen ‘and then’ as a linker of turn continuations in L1-speakers, we cannot see any differences in the interactional deployment among L2 speakers. Both L1 and L2 speakers use å sen to link continuations post gap, post other talk and in the next turn. The discursive motivations seem to be similar: to give a new direction to what the speaker’s just completed turn has presented or, as the next speaker, to change the footing and introduce a thematically linked but a more personal, and thus potentially non-aligning, continuation to the collective interactional project. In other words, the L2-speakers have learnt and internalized the use of an interactional device in their second language (see e.g. Berger and Pekarek Doehler, 2018 and Skogmyr Marian, 2020 for L2 interactional competence).

However, there are some L2-specific features of how å sen is deployed grammatically. The syntactic norm is that the adverbial item sen ‘then’ triggers subject–verb inversion in accordance with the basic Germanic verb-second constraint in declarative sentences (cf. Dutch and German). This means that å sen operates within the clausal frame (although the conjunction å ‘and’ does not). Some of the L2-speakers in the data treat å sen as an item outside of the clausal frame, in the so-called “pre-front field” (see Auer, 1996b), which results in a non-canonical V3-structure, i.e. adverbial–subject–verb. This happens in five instances out of the total of 24 continuations, which means that it is not a dominant pattern, but is nonetheless a pattern that has been identified as a typical L2 feature in previous research (Viberg, 1992) but also as a characteristic of contemporary Swedish urban vernaculars (Ganuza, 2010). Such syntactic variation does not seem to depend only on the speaker’s grammatical skills, however. We can consider an enlightening part of Excerpt 3 below. In line 11, Tanja initiates a clause with the adverb därför ‘therefore,’ which correctly triggers inversion in her talk. Interestingly, in line 17, she introduces a turn continuation with å sen that is followed by a clause with a straight, non-inverted word order.

Apparently, the speaker masters the V2-rule as well as the fairly advanced placement of the negation before the finite verb in a dependent clause, as seen in line 13 (så att man inte kunde se, lit. ‘so that you not could see’). The deployment of å sen outside of the clausal frame in line 17 suggests that this compound connective is treated on a par with ordinary conjunctions and discourse markers. It is followed by a micropause, which reflects its production as a chunk and a unit of its own, but also as a prominent element that organizes progressivity in the speaker’s description. Similar uses of the adverbial (å) sen in contemporary urban vernaculars have been attributed to a more liberal realization of information-structural strategies than in standard varieties (Freywald et al., 2015). Furthermore, it is worth considering that å sen, produced as a chunk and followed by a micropause above, offers the speaker a light-weight turn-entry device (see Sacks et al., 1974: 719), which is one of the functional motivations of pre-front-field elements and favors early starters (see Auer, 1996b). When the right to the turn space has been established with an opening, the speaker can start to organize the core of the contribution from a clean slate, so to speak, as suggested by the straight word order following å sen above.

Indeed, something similar occasionally happens in L1 talk. We have two cases in our L1 data in which å sen does not trigger inversion, one of them presented in Excerpt 6, line 11. The painting under discussion is Frida Kahlo’s A Few Little Nips, which has some unconventional solutions concerning perspective.

In lines 1 and 3, Magnus is finding out the name of the painting, first reading it in the background material about the artwork, and then he deduces that the name is also spelled out in Spanish in the artwork itself. Sabina ratifies this observation rather emphatically in line 6, and then she possibly marks sequence closure with the acknowledgment token mm in line 9. In line 11 she initiates a post-gap continuation with å sen. This continuation diverges considerably from what the prior sequence was about, i.e. Magnus’s reference to the name of the painting. Instead, Sabina continues to talk about the artist’s use of perspective. The deployment of å sen outside of the clausal frame, i.e. as a typical discourse marker, may be connected to this greater discursive shift: there is a sequence boundary and Sabina’s continuation does not address any further what Magnus had pointed out but continues with something that is a personal issue, presented as a “pondering” (ja funderar “I wonder”).

It is also of interest that Sabina is very fluent and the most talkative of the participants in the conversation, having quite a fast speech rate. She regularly deploys different kinds of light turn and clause entries that are abandoned and followed by a restart with a new clausal unit. All this enables her to inhabit the turn space and produce more language. We can see several of these phenomena in the internal organization of Sabina’s turn in lines 12–13: varför har konstnären valt- ‘why has the artist chosen-’; de e ju liksom, ‘it’s like you know,’; domhä perspektiven e ganska, ‘these perspectives are quite.’ The use of å sen in the pre-front field fits in this pattern; it offers the speaker a turn entry with a weak projecting force, leaving the syntactic gestalt opening underdetermined (see Auer, 1996b: 297). Moreover, the complement-taking predicate (see Thompson, 2002) ja funderar following å sen at the beginning of Sabina’s turn in line 11 often appears in the pre-front field in spoken Swedish (see Lindström and Lindholm, 2009). Sabina’s turn beginning reflects such a spoken-language use: ja funderar också ‘I’m pondering also’ is followed by the complementizer att and an interrogative with main clause syntax, instead of just an embedded interrogative as the object. Thus, Sabina seems to construct her turn entry with a series of discourse-marker like fragments, possibly to patch up the thematic-sequential boundary to the previous speaker’s contribution.

This observed syntactic variability in L1 talk suggests that å sen may be practiced strategically in certain discourse contexts and the dynamics of turn-taking. This pattern of use may be in a somewhat more liberal use in L2 talk, but not only as a feature of a language learner’s speech but with a functional motivation that can be activated when necessary, as has also been observed in speakers of contemporary urban vernaculars. As a compound unit with the conjunction å ‘and,’ å sen can be on a grammaticalization cline toward the category of discourse markers (where å belongs) that in syntactic terms inhabit the pre-front field.

As for the embodied conduct, there are some converging and diverging tendencies in the practices of L1 and L2 speakers. The embodied pattern is similar when it comes to how the speakers construct a turn as current, completed and recompleted: gesticulation is active during a current turn and a turn continuation, while gesture retraction to a rest position occurs at possible turn completion. This is perhaps not so surprising in light of previous research that has established this kind of embodied orientation in speakers of different languages and cultures (Sacks and Schegloff, 2002; Li, 2013; Mondada, 2015), i.e. an embodied sensitivity to a turn’s status as current or completed may be an interactional universal. However, we can see a more prominent L2 feature in the occasionally heavily used depicting hand gestures that can also relate to the speaker’s own body parts as a reference point (see Excerpt 3). These depicting gestures accompany the verbal language and may be felt as a useful support when the speaker is not completely reliant on the mastery of a specific vocabulary.